Peceneaga-Camena Fault: Geomagnetic insights into active tectonic contact

Highly detailed, very accurate ground magnetic investigations were jointly conducted by Romanian and Ukrainian researchers on a segment of the Peceneaga-Camenas Fault (PCF) in order to reveal the potential of geomagnetic method for active faults investigating. The survey succeeded to outline the PCF track in the area covered by recent sediments, and provide insights on the fault structure and in-depth development. 2 numerical modeling has been employed for interpreting the obtained geomagnetic anomaly. Lateral variations in magnetization, as suggested by the model, reveal the complex geological architecture in the area, hidden by recent deposits. The zero magnetization outlined in the central part of the survey lines has been interpreted in geodynamic terms, as a breccias zone created along PCF track by its active dynamics.

General consideration. The Peceneaga-Camena Fault (PCF) represents one of the most studied tectonic features in the Romanian territory, even from the beginning of the 20 th century [Mrazec, 1912;Macovei, 1912]. It generally appears (Fig. 1) as the boundary between the Moesian Platform (MP), represented in the area by Central Dobrogea ( CD ) , and North Dobrogea (ND) geological units.
Geophysics brought significant evidence on the PCF in-depth extent. The international deep seismic soundings (DSS) line ¹ 2 [Radulescu et al., 1976 ] has revealed its crustal nature , showing a step of about 10 km at the both Conrad and Moho discontinuities. Later on, seismic tomography images based on CALIXTO experiment [Martin et al., 2006], have revealed PCF as a major lithospheric contact between East European Plate (EEP) and Moesian Micro-plate (MoP) reactivated during the W Black Sea opening [Besutiu, Zugravescu, 2004;Besutiu, 2009].
The Baspunar Geodynamic Observatory (BGD) was especially designed and run by the Solid Earth Dynamics Department at the Institute of Geodynamics of the Romanian Academy in order to moni- Fig. 1. Simplified tectonic setting of PCF and location of the study area: 1 -North Dobrogea boundaries (acropping out, b -covered); 2 -strike-slip faults; 3 -structural axes (a -syncline, b -anticline); 4boundaries between North Dobrogea main units (a -cropping out; b -buried); 5 -Cirjelari-Camena Outcrop Belt (a -cropping out, b -covered); 6 -episutural post-tectonic cover; 7 -river; 8 -settlements (a -major cities; b -villages); 9 -Baspunar Geodynamic Observatory (BGD) location; PDD -Predobrogean Depression; ND -North Dobrogea; CD -Central Dobrogea; BB -Babadag Basin. tor slip along PCF. This paper mainly deals with results of the high accuracy detailed magnetic investigations carried out on the PCF segment located in the neighborhood of the BGD, aimed at revealing the path and in-depth structure of the PCF in the monitoring area. Research has been carried out in the frame of the bi-lateral project INRAF ("Integrated research of some active faults located in the NW inland of the Black Sea on the Romanian and Ukrainian territories"), jointly developed by the Institute of Geodynamics of the Romanian Academy and the Institute of Geophysics of the National Academy of Sciences of Ukraine.
The local geological background. North Dobrogea. The area subject to geophysical investigation mainly belongs to the so called Cirjelari-Ca-mena Outcrop Belt (CCOB). A thorough description of the structure and lithostratigraphy of this unit was provided by Gradinaru [Gradinaru 1980[Gradinaru , 1984[Gradinaru , 1988, and a simplified geological sketch for the study area is shown in Fig. 2, along with the location of the magnetically surveyed panels.
On the overall, the study area is dominated by the presence of the Jurassic sedimentary and volcanic rocks, unconformable overlying older Palaeozoic deposits of the Macin Unit and largely covered by the post-tectonic sedimentary cover of the Cretaceous Babadag Basin and shallow Quaternary formations.
The Quaternary rocks are mainly represented by shallow layers of loess deposits.
The Palaeozoic formations are practically hidden by younger deposits, except for some confined areas where they may crop out (Cirjelari valley, where Aiorman Fm unconformable lies on Palaeozoic basement rocks).
The Baspunar spilite (J 3 ) is mainly represented by pillow-lava flows interbeded in Jurassic limestone.
Since early times, it has been also noticed [Motas, 1913] that CD deposits in the PCF contact zone are sometimes intercalated with, or intruded by magmatic rocks of North Dobrogea (rhyolites), which might represent another source for geomagnetic anomalies.
Data acquisition and processing. Field observations were conducted by using two G 856 AX magnetometers (one for the record of diurnal geomagnetic activity, and the second one for observations along the survey lines). Basically, the survey lines were designed almost perpendicular to the assumed PCF track. The lines are 4 m apart, and a step of 2 m between two consecutive stations along each line was used to survey the study area. Location of data points was set by using a Garmin 78 GPS receiver. The geographic coordinates on WGS 1984 ellipsoid were then transferred into the rectangular coordinates  of the Romanian national stereographic projection system [Avramiuc et al., 2001].
The geomagnetic sensor was placed at 3 m above the ground in order to avoid (or at least to mitigate) shallow local effects.
Diurnal geomagnetic activity was observed and recorded every minute during the survey in a local base-station, located close to the surveyed area.
Routine processing has been applied to the raw observations in order to provide data consistency: removal of the effect of external sources and base reduction.
As a result, a time-invariant ∆T as referred to the survey base-station was obtained. Finally, a residual geomagnetic anomaly was computed by removing a first-order polynomial trend from the observations, and ∆T a geomagnetic maps were plotted (Fig. 3).
Modeling geomagnetic sources. Taking into consideration the pattern of the geomagnetic anomaly and some previous information on the study area, attempts for modeling the geomagnetic sources and their geological interpretation have been performed. The software. The professional GM-SYS® software run on the Geosoft OASIS® platform has been used for 2D modeling along the survey lines. It is based on the methods proposed by [Talwani et al., 1959;Talwani, Heirtzler, 1964], and employs algorithms published by [Won, Bevis, 1987]. Rocks magnetic properties. Magnetic properties of the rocks in the area have been conside-red according to previous rock physics determinations [Besutiu, 1997;Besutiu, Nicolescu, 1999], to which additional determinations on outcrops samples were performed in the IG-NASU laboratory. Table 1 shows some magnetic properties of the main geological formations within North Dobrogea and the study area.
Polarizing field. As Köenigsberger coefficient (Q) of the geological formations known in the study area generally shows small values, the induced magnetization model has been considered during the computation, with the following parameters for the polarising field: -total intensity field T = 48 500 nT, -geomagnetic inclination I = 62° N, -geomagnetic declination D = 3° E. Physical model. Two main aspects of the modeling should be stressed: -to get a better fit between the observed and predicted anomaly, laterally extended geomagnetic source models (exceeding the survey line) were taken into consideration; -due to the close vicinity of the survey lines, similar geomagnetic patterns, and, consequently, except for small lateral changes in geometry, rather similar models of the sources of the geomagnetic anomalies as showed in the followings were outlined along various lines. Basically, following the trial & error process of 2D modeling along the survey lines the best fit has been obtained for the geometry and rock magnetic properties as illustrated in Fig. 4. 1 -observed field, 2 -predicted effect, 3 -magnetic body ID (see Table 2). Overall, to predict the geomagnetic anomaly, several 2D magnetic bodies have been considered. Table 2 shows their magnetic susceptibilities along with the assumed geological significance.
Geological interpretation. Based on previously gathered tectonic knowledge and rock physics of the main geological formations occurring in the study area and neighbouring region, an attempt for interpreting the geomagnetic sources outlined by modeling has been made. The results are synthetically illustrated in Fig. 5. As previously mentioned, the interpretative geological cross-section laterally extends over the magnetic line in order to mitigate the effect of the signal truncation and side effects.
Overall, the geological interpretation of the synthetic model has allowed outlining the PCF path by separating PCF flanks due to the general distinct geomagnetic behaviour of their different embedded geological formations (basically magnetic CD Proterozoic GSS versus non-magnetic ND Palaeozoic sedimentary).
But, the survey accuracy has also allowed discriminating some distinct layers with different magnetization within GSS, as well as the presence of some intrusive rocks (diorite dykes?) penetrating the geological formations.
On the other hand, basalt flows (Baºpunar spilite) embedded within the Baºpunar Fm (Jurassic and/or Triassic limestone) significantly complicate T a b l e 2 . Magnetic properties of the source model the interpretation by locally increasing the geomagnetic behaviour of the respective sedimentary pile.
Revealing the PCF track. Taking into consideration the peculiarities of the geomagnetic field pattern over the two flanks of the PCF, and the results of the quantitative interpretation of the 2D modeling along the survey lines, PCF track could be clearly outlined in the areas covered by recent deposits except for some confined areas due to the insufficient westward extension of the survey lines Fig. 6. Geodynamic considerations. One interesting aspect pointed out by the geomagnetic modeling has been the apparent lack of magnetic properties in the central compartment of the interpreta-tive cross-section, located along the assumed PCF track. This has been interpreted in terms of fragmentation of the PCF flanks, generating rock-debris through the abrasion of the fault flanks as a consequence of its active slip. Despite some initial individual magnetic properties, on the overall, elements of this compartment may not be reflected in the pattern of the geomagnetic anomaly due to the current of randomly distributed direction of magnetization of the breccias elements. Besides, water circulating within the contact zone accelerated the magnetic minerals weathering and, consequently, the loss/mitigation of original magnetic properties. Concluding remarks. Detailed high accuracy ground magnetic survey on a PCF segment located in the vicinity of BGD succeeded to outline the fault track and in-depth structure, based on the interpretation of some 2D models simulating sources of geomagnetic effects. The active character of the fault has been indirectly revealed through the loose of magnetic be-